Agent intelligence through data mining. (English) Zbl 1106.68099

Multiagent Systems, Artificial Societies, and Simulated Organizations 14. New York, NY: Springer (ISBN 0-387-24352-6/hbk). xxv, 201 p. (2005).
This book is a result of research experience of the authors with the “Agent Academy” frame, which is a three-years-long project funded by the European Commission. The theme of this book is to develop an integrated framework for constructing multi-agent applications and for improving agent intelligence by the exploitation of data mining techniques. It addresses a challenging problem of generating intelligence from data and transferring to a separate, possibly autonomous, software agent. Several existing data mining techniques are considered and some new ones are described.
This book starts with a brief introduction, reviews on major key techniques on data mining as well as intelligent agents, and presents the methodology that essentially embeds knowledge models, extracted via data mining, into agents and multi-agent systems. The methodology is applied to three distinct levels of knowledge acquisition: the application data level, the agent-behavior or data level, and agent communities. One third of this book is devoted to describe the three representative test cases where the readers would see how the developed theories are really implemented.


68T05 Learning and adaptive systems in artificial intelligence
68T35 Theory of languages and software systems (knowledge-based systems, expert systems, etc.) for artificial intelligence
68P15 Database theory
68-02 Research exposition (monographs, survey articles) pertaining to computer science